\input epsf
\nopagenumbers
\magnification=\magstep1

\noindent {\bf Problem statement} Each of four real numbers is known
to lie between 0 and 50. The product of these four numbers is
approximated by first rounding each number to the nearest tenth (that
is, to the nearest number with one digit after the decimal point) and
then taking the product of these rounded numbers. Use differentials
(linear approximation) to estimate the maximum possible error in the
approximate product.



\vfil\eject\end